Skip to content

status.openkey.ai is live — independent, worldwide uptime proof See it now ->

openkey

Mistral AI

19 models built · 17 served

Mistral AI is a French lab founded in 2023, shipping both open-weight and commercial models across the Mistral Large, Mistral Small, and Ministral families. Through OpenKey, 19 Mistral models are available, ranging from the 3B-parameter Ministral 3 3B up to Mistral Medium 3.5. The lineup covers general text and vision models plus specialists: Codestral and Devstral for code, Voxtral for audio-to-text, and Saba for lighter-weight tasks. Pricing spans from $0.02 per 1M input tokens (Mistral Nemo) up to $2.00 per 1M input tokens (Mistral Large, Mixtral 8x22B). None of Mistral's models on OpenKey are free — this is a commercial lineup with tiers for budget, mid-range, and flagship use.

Lineup

For general-purpose work, Mistral Medium 3.5 ($1.50/$7.50 per 1M in/out) is the current flagship, with a 262,144-token context window and text+image+file input. If you're budget-constrained, Mistral Small 4 ($0.15/$0.60) or Mistral Small 3.2 24B ($0.075/$0.20) cover most everyday text and vision tasks at a fraction of the cost. For code, use Devstral 2 ($0.40/$2.00) or Codestral 2508 ($0.30/$0.90). For the cheapest option in the whole lineup, Mistral Nemo runs $0.02/$0.03 per 1M tokens — legacy pricing, but still live. Ministral 3B/8B/14B models ($0.10-$0.20) suit lightweight, latency-sensitive jobs.

Prices per 1M tokens, flat 3% fee included.

Models Mistral AI serves

Beyond its own models, Mistral AI hosts 17 models in the OpenKey catalog. Prices are Mistral AI’s own list price per 1M input tokens, before our fee.

From the OpenRouter catalog, Jul 28, 2026. All providers

Status page ↗Privacy policy ↗Terms of service ↗

Questions

What's the price range for Mistral AI models on OpenKey?
Input pricing runs from $0.02 per 1M tokens (Mistral Nemo) to $2.00 per 1M tokens (Mistral Large, Mistral Large 2407, Mixtral 8x22B). Output pricing runs from $0.03 to $7.50 per 1M tokens. OpenKey charges provider list price plus a flat 3% fee — for Mistral Medium 3.5, that's $1.50 x 1.03 = $1.545 input and $7.50 x 1.03 = $7.725 output per 1M tokens.
Which Mistral model is cheapest, and which is the best?
Mistral Nemo is the cheapest at $0.02/$0.03 per 1M input/output tokens. Mistral Medium 3.5, released April 2026, is the newest and most capable general model, with a 262,144-token context window and text+image+file input support. There are no free Mistral models on OpenKey — all 19 are paid.
How do I access Mistral AI models through OpenKey?
All 19 Mistral models — including Mistral Large, Mistral Small, Ministral, Codestral, Devstral, Voxtral, and Saba — are available under one OpenKey API key, alongside 329 other models from 52 labs. No separate Mistral account or billing setup needed; pricing is provider list price plus a flat 3% fee.